[0001] This invention relates to a thermite process for producing a metal or alloy and,
more particularly, it relates to a process for efficiently and effectively producing
high quality ferroniobium or metallic chromium by an ingenious technique of charging
the furnace with a mixture of starting materials.
[0002] A thermite process has been known for long as a method of producing a metal or alloy
by reducing an oxide of the metal or an ore containing an oxide of the metal by another
metal such as aluminum or silicon in a powdered state.
[0003] The thermite process can be applied for production of a variety of metals and alloys
including metallic chromium, ferrovanadium, ferroniobium, ferroboron and other ferroalloys.
[0004] When any of these alloys or metals is prepared by a conventional thermite process,
the starting metallic oxide or ore containing the metallic oxide is crushed and mixed
with a reducing agent such as powdered aluminum, to which a slag forming material,
an exothermic agent and/or a cooling agent are added if necessary. The mixture is
then put in an appropriate furnace and ignited for thermite reaction. The reaction
of the mixed materials proceeds exothermically to produce a metal or alloy and a slag
under a fused condition, which are then cooled, solidified and separated from each
other to obtain the intended metal or alloy.
[0005] For producing a metal or alloy by a known thermite process as described above, it
is important to keep the temperature of the fused materials formed by at or above
2,000°C in order to effectively separate the intended metal or alloy from the fused
slag. However, since the temperature of the thermite reaction is solely dependent
on the heat generated by the exothermic thermite reaction, which can significantly
vary as a function of the metal involved, it may be too high or too low depending
on the type of the metal or alloy to be produced.
[0006] If, for example, the amount of the heat generated by the reaction is too large, the
process proceeds too vehemently to spatter the raw materials and consequently lower
the yield. Besides, it may severely erode the lining of the furnace and the security
may be threatened. If, on the other hand, the amount of the heat is too small, the
yield will also be lowered because the reaction does not proceed at a satisfactory
rate and the resultant metal or alloy may contain some of the slag to deteriorate
its quality. If the amount of the generated heat is exceptionally too small, the reaction
can terminate while the starting materials are only partly fused.
[0007] Therefore, it is essential for a thermite process to ensure an appropriate level
of heat value, particularly when a metal or alloy is to be produced on an industrial
basis and there have been taken specifically designed measures to meet this requirement.
Some of the measures include the following.
(1) For an excessively exothermic reaction, the heat generated in the reaction is
suppressed by partly replacing the oxide of the metal in the reaction system with
the pure metal. Alternatively, a volume of an undersized material of the metal or
alloy to be produced (that simply consumes heat to become fused and does not generate
heat) is added to suppress the temperature.
(2) For an insufficiently exothermic reaction, an electric furnace is used to supply
additional heat to the reaction system. Alternatively, a mixture of powdered aluminum
and an easily decomposable oxide that scarcely consumes heat for decomposition is
added to the reaction system as an exothermic agent.
[0008] The operation of a known ordinary thermite process is conducted on a batch basis.
A batch of the starting materials are weighed, mixed and then loaded into a furnace
in one lot at a time, the amount of the heat to be generated in the reaction being
determined by calculating the average exothermic energy of the batch in the reaction
system. Of course, meticulous calculations and preliminary experiments are required
for determining the exact amount of the starting materials.
[0009] However, such a known process is accompanied by problems as described below. Firstly,
according to a close observation by the inventors of the present invention on the
metallurgical reaction in a known thermite process conducted on a batch basis, the
reaction proceeds very slowly in the initial stages, revealing that a heat-insufficiency
condition is there, followed by the intermediary stages where the reaction is suddenly
accelerated until an excessively high rate of heat generation emerges. When the thermite
reaction proceeds with such a profile, there can be cases where the reaction comes
to a stand-still somewhere in the initial stages and those where the starting materials
are spattered about to lower the yield in the latter part of the reaction while the
lining of the furnace becomes very liable to be eroded. Besides, the reaction can
proceed very irregularly with the known process so that the resultant metal or alloy
may be of inferior quality.
[0010] It is therefore the object of the present invention to provide a thermite process
that can overcome the above described problems of a known process which are attributable
to the fact that the process is conducted on a batch basis and consequently produce
a high quality metal or alloy at a high yield.
[0011] As a result of intensive research efforts to achieve the above object, the inventors
of the present invention came to find a new thermite process with which the heat generated
in a thermite reaction can be regulated by appropriately using cooling and exothermic
agents at rates that can be controlled at each stage of the metallurgic reaction so
that a high quality metal or alloy may be produced at a high yield.
[0012] According to the invention, there is provided a thermite process for producing a
metal or alloy by charging a reactor furnace with a mixture of a powdered metallic
oxide and a reducing agent such as powdered aluminum, wherein a batch of the mixture
of the starting materials is divided into a plurality of loading lots in such a manner
that the amount of heat generated by each of the loading lots of the starting materials
is so regulated that it differs from any of the others and the loading lots are sequentially
arranged and loaded into a furnace for thermite reaction in an ascending order in
terms of the level of heat generation. The regulation of the heat generation in each
of the loading lots of the starting materials is realized by using a stoichiometric
mixture of at least one of sodium chlorate, potassium chlorate, potassium perchlorate,
sodium nitrate, potassium nitrate, calcium peroxide, barium peroxide and chromate
and a reducing agent such as aluminum and by varying the amount of the mixture added
to each of the loading lot. Preferably, the powdered metallic oxide is chromium oxide
or iron oxide for metallic chromium and niobium oxide for ferroniobium.
[0013] In the following, the present invention will be described in greater detail by way
of examples.
[0014] The reaction for producing metallic chromium from, for instance, chromium (III) oxide
(Cr₂O₃) and aluminum (Al) by way of a thermite process will be expressed by the following
formula.
Cr₂O₃ + 2Al = 2Cr + Al₂O₃ [1]
[0015] According to the above formula, the aluminum oxide (Al₂O₃) corresponding to the amount
of metallic chromium is produced.
[0016] When the starting materials are compounded for a thermite reaction for producing
a metal or alloy by using aluminum as a reducing agent, the amount of aluminum to
be charged is a function of the kind and the composition of the metal or alloy to
be produced. For instance, if the metal or alloy to be produced may contain aluminum
to some extent, the amount of aluminum to be used is made somewhat larger than the
stoichiometric quantity determined by the applicable chemical formula in order to
improve the yield. This is a way normally applied for producing ferroalloys. If, on
the other hand, the metal to be produced should be free from aluminum as much as possible,
as in the case of metallic chromium, the amount of aluminum to be used is made somewhat
smaller than the stoichiometric quantity so that the aluminum may be entirely converted
into aluminum oxide. This is a way that has long been practiced in the production
of metallic chromium and there are a number of papers on this theme. For instance,
"ANNALI DI CHIMICA" 77 ∼ 81, Vol. 41 (1991), "Rassian Metallurgy and Mining" 20 ∼
26, No. 1 (1963), "Ullmanns Encyklopadie der Technischen Chemie" Band 9, 591, (1975)
etc.
[0017] According to these papers, the amount of aluminum to be used is normally less than
the stoichiometric quantity by 5 to 20 percents.
[0018] When the amount of aluminum to be used is smaller than the stoichiometric quantity,
some of the chromium (III) oxide charged into the furnace is inevitably left unaffected
in the reaction system after completion of the reaction to form slag with aluminum
oxide, which is the principal component of the slag. If, for instance, the amount
of aluminum charged into the furnace is 90% of the stoichiometric quantity, the produced
slag will contain about 85% of Al₂O₃ and about 15% of Cr₂O₃. While the slag to metallic
chromium ratio will be substantially equal to one (1) when aluminum is used by a stoichiometric
quantity, it will become greater as the amount of aluminum is reduced in the reaction
system because the amount of Cr₂O₃ which is left unaffected is increased and the amount
of metallic chromium produced goes down, making the slag to metallic chromium ratio
consequently go up. If the amount of aluminum is 90% of the stoichiometric quantity,
the slag to chromium ratio will be 1.15.
[0019] Molten metallic chromium and slag containing mainly Al₂O₃ are simultaneously produced
in a mixed state as the reaction proceeds and then separated from each other by the
difference of specific gravity. Separation of metallic chromium slag should be conducted
quickly and as soon as possible after the production of mixture of metallic chromium
and slag in order to obtain the metallic chromium of high purity concentration.
[0020] For effective separation of metallic chromium and slag, temperature plays the most
important role. If the generation of heat is small and the temperature is low, the
slag will not be satisfactorily separated from the metallic chromium. On the other
hand, the reaction violently proceeds within a short period of time if a high heating
value is involved so that some of the produced molten metal can precipitate in the
reaction furnace and is cooled and solidified there before it is satisfactorily separated
from the slag while the starting materials may partly remain unaffected because the
reaction time is too short. In order to complete the reaction and effectively separate
the produced metal or alloy from the slag, two contradictory requirements of a high
temperature for completion of reaction and a long reaction time for effective separation
of metal and slag should be simultaneously met. These requirements can be met only
through elaborate control of heat generation.
[0021] If the process of thermite reaction for producing a metal or alloy is carefully observed,
it will be understood that the reaction proceeds in the furnace in the following manner.
[0022] After loading the starting materials into a reaction furnace, a thermite reaction
is started by igniting the top of the materials (e.g., a mixture of a powdered metallic
oxide and powdered aluminum) loaded in the furnace. The materials are gradually fused
by the heat generated in the exothermic thermite reaction to become a molten slag,
which heat the starting materials in the neighboring area to cause them to react with
each other. The reaction proceeds in this way until all the materials are consumed,
when the reaction finally terminates. Since the quantity of the molten slag is relatively
small and that of the materials left for future reaction is large in the initial stages,
heat is always in short supply for the reaction system in those stages, to make the
reaction proceed slowly. In other words, the thermite reaction proceeds only along
the interface of the loaded starting materials and the molten slag, and the interface
gradually goes down. When the reaction has advanced to a certain degree and the interface
of the loaded starting materials and the molten slag has a relatively large area while
it stores a considerable amount of heat in it, the reaction can become very vehement.
However, such as uneven progress of reaction is not desirable for maintaining a desirable
temperature and achieving a good separation of the metal and slag. Theoretically,
any undesirable condition of generating heat may be checked by regulating the rate
of heat generation in every stage of the reaction. However, an attempt to constantly
regulate the rate of heat generation in the reaction system will not be successful
because of the high temperature of the molten slag, which is somewhere around 2,000°C.
One possible method of regulating the rate of heat generation may be preparing a small
amount of molten slag in the reactor furnace and gradually introducing a corresponding
amount of the starting materials into the furnace. However, this method is not feasible
because the materials placed on the slag is vehemently spattered about to seriously
lower the yield of the final product.
[0023] In view of these circumstances and as a result of intensive research efforts, the
inventors of the present invention came to find that the above described problems
can be effectively solved by dividing a batch of a mixture of the starting materials
into several loading lots in such a manner that the amount of heat generated by each
of the lots is different from those of the other lots and loading the lots sequentially
into the furnace in an ascending order in terms of the level of heat generation. With
such an arrangement of preparing a plurality of loading lots out of a batch of starting
materials for regulating the heat generated in a thermite reaction, the above described
problems are effectively solved.
[0024] A metallic to oxide be used as one of the primary starting materials may be, if the
intended product is metallic chromium or a ferroalloy, iron oxide, chromium oxide,
manganese oxide, niobium oxide, vanadium oxide, molybdenum oxide, boron oxide, titanium
oxide, tungsten oxide or an one containing any of these. When the material is supplied
in the form of blocks, they should be crushed in advance to granules of an appropriate
size, preferably 100 mesh or less. If the size of the particles is too large, the
reduction by aluminum does not proceed at a sufficient rate and a prolonged reaction
time can result in an insufficient supply of heat and a low yield of the intended
metal or alloy.
[0025] Aluminum to be used as a reducing agent is normally supplied in the form of granules,
flakes or needle-shaped pieces having a size of approximately 3 mm. It may be partly
replaced by another reducing agent such as silicon, calcium or magnesium.
[0026] According to the invention, a slag forming material is used as an auxiliary starting
material which is added to said primary starting materials. When a metallic oxide
and aluminum are mixed and put to thermite reaction in a furnace, the reaction products
in the furnace will be the metal or alloy and a slag almost entirely consisted of
alumina. Since alumina melts only at very high temperature, measures should be taken
to lower the melting point for the purpose of the thermite reaction and the slag forming
material is used exactly to lower the melting point of the slag. The slag forming
material according to the invention will be quick lime, fluorite, magnesia or any
other appropriate material.
[0027] A cooling agent may be used to regulate the rate of heat generation. Such a cooling
agent should not interfere with the thermite reaction simply melt to absorb excessive
heat in the reaction system. Normally, sieved particles of the intended metal or alloy
will be used as they may be advantageously remelted. Since the slag forming material
used for the reaction system does not interfere with the thermite reaction either,
it may also act as a cooling agent.
[0028] Another auxiliary material will be an exothermic agent. This is a material to be
used when the rate of heat generation in the reaction system is too low and normally
a stoichiometric mixture of an oxygen supplying material and alumina. In most cases,
the oxygen supplying material will be a peroxide. It should not affect the quality
of the intended metal or alloy and the chemical properties of the slag and may be
preferably selected from sodium chlorate, potassium chlorate, sodium perchlorate,
potassium perchlorate, sodium nitrate, potassium nitrate, calcium peroxide and barium
peroxide. Chromates may also be suitably used when the intended metal or alloy is
metallic chromium or a chromium alloy.
[0029] As described earlier, a thermite process according to the present invention is characterized
in that the above mentioned starting materials are mixed to form a batch, which is
divided into a plurality of loading lots in such a manner that amount of heat generated
by each of the loading lots of the starting materials differs from any of the others.
For instance, a batch of a metal oxide to be subjected to thermite reaction may have
a weight of 1,000 kg and it may be divided into four lots, each weighting 250 kg,
although the amount of heat each of the lots generates differs from any of the others
and is so adjusted as to best fit the location in the furnace where it is loaded.
[0030] The amount of heat to be generated by each of the lots of the mixture of the primary
starting materials is regulated by determining the amount of the slag forming material
and that of the cooling material to be added to the mixture of the metallic oxide
and aluminum having a predetermined weight. The amount of an exothermic agent, e.g.,
a stoichiometry mixture of sodium chlorate and aluminum, to be added to the mixture
of the primary starting materials is further determined so that the amount of heat
generated by the lot of the final mixture exactly meets the requirement of heat generation
which is a function of the location of the furnace where the lot is loaded.
[0031] The total amount of heat required to carry out the thermite reaction of a batch solely
depends on the type of the intended metal or alloy. In the case of producing ferroniobium
starting from niobium oxide ore and iron oxide, the heat requirement will be approximately
600 kcal/kg. For producing metallic chromium from chromium (III) oxide, the required
rate of heat supply will be about 720 kcal/kg. For producing ferrovanadium from vanadium
pentaoxide and iron oxide, heat will need to be supplied at a rate of about 650 kcal/kg,
whereas, in the case of producing ferromolybdenum from molybdenum trioxide and iron
oxide, the rate will be about 560 kcal/kg.
[0032] On the other hand, the amount of heat required to carry out the thermite reaction
of each of the lots obtained by dividing a batch is a function of the type of the
intended metal or alloy, the particle size of each of the starting materials, the
profile of the furnace to be used, the amount of the batch and other variables. Therefore,
the rate of addition of an exothermic agent should be exactly determined through experiments.
According to the result of a research conducted by the inventors of the present invention,
the amount of heat generation of a first lot of a batch for producing ferroniobium
always fall short of heat, although there is supplementary supply of heat by an exothermic
agent (See Table 2 in Example 1).
[0033] The lots of a batch obtained by weighing and mixing the starting materials are then
sequentially loaded into the reactor furnace in an ascending order in terms of the
level of heat generation. The reactor furnace may of any appropriate profile, although
a standing cylindrical furnace is preferable as the lets are loaded in so many layers
and, therefore, the furnace should have a limited horizontal sectional area. Most
preferably, when a batch of the starting materials is completely loaded into the furnace,
it shows a height/diameter ratio of 1 to 1.5.
[0034] After the total loading lots of the starting materials is loaded into the reactor
furnace, it is ignited for thermite reaction. While the time required for the reaction
to complete depends on the type of the intended metal or alloy, it is desirable that
the reaction proceeds at a constant rate. If any fluctuation in the rate of reaction
is observed in any stage of the reaction, the rate of heat generation should be regulated
by the next batch to be loaded into the furnace.
[0035] After completion of the reaction, the fused metal or alloy and the slag are separated
from each other due to the difference of specific gravity and the former is deposited
on the bottom of the reactor furnace. It can be collected after it has been cooled.
[0036] With a thermite process according to the invention, since the temperature of the
reaction product can be appropriately maintained after the reaction is completed,
the obtained fused metal or alloy and the fused slag can be separated from each other
very easily and the former is of very high quality as it scarcely contains non-metallic
inclusion and other impurities. Moreover, since the starting materials are hardly
spattered around, the yield will be exceptionally high.
[Example 1]
[0037] Ferroniobium was produced, using niobium ore (niobium oxide) and iron oxide as starting
metallic oxides, aluminum as a reducing agent, sodium chlorate as an exothermic agent
and a mixture of fluorite and quick lime as a slag forming material. Table 1 shows
the composition and the particle size of these materials.
[0038] A standing cylindrical reactor furnace having an inner diameter of 1 m and a height
of 2 m was used. A batch of the materials contained a 1,000 kg of niobium ore and
was divided into three lots. Each of the first and second lots contained a 400 kg
of niobium ore, while the remaining portion, or 200 kg, of the niobium ore, was contained
in the third lot. Each of the other component materials was also divided into three
portions and the materials were mixed together for each lot with such a ratio that
each of the lots would generate an intended amount of heat. The ratio of the component
materials and the rate of heat generation for each of the three lots are shown in
Table 2.
[0039] For thermite reaction, the first lot was loaded into the reactor furnace and the
second lot was placed on the first lot, the third lot being subsequently placed on
the second lot. The top of the load was ignited to trigger a thermite reaction. The
reaction went on evenly until it was terminated twelve minutes after the start of
the reaction. When the reaction product was cooled, the slag found on the upper portion
of the product was removed and the ferroniobium, remained in the furnace was collected.
Table 1
|
Composition (%) |
Particle size (mm) |
|
Nb₂O₅ |
Ta₂O₅ |
FeO |
SiO₂ |
CaO |
CaF₂ |
|
niobium ore |
60.1 |
0.4 |
5.2 |
2.9 |
12.9 |
-- |
0.15 or less |
iron oxide |
-- |
-- |
73.0 |
-- |
-- |
-- |
10 or less |
quick lime |
-- |
-- |
-- |
-- |
94.0 |
-- |
10 or less |
fluorite |
-- |
-- |
-- |
0.8 |
-- |
96.0 |
5 or less |
aluminum |
Al : 97.3 |
3 or less |
sodium chlorate |
NaClO₃ : 98.0 |
1 or less |
Table 2
|
1st lot |
2nd lot |
3rd lot |
total |
niobium ore (kg) |
400 |
400 |
200 |
1000 |
iron oxide (kg) |
65 |
65 |
32 |
162 |
aluminum (kg) |
60 |
160 |
130 |
350 |
sodium chlorate (kg) |
17 |
20 |
13 |
50 |
quick lime (kg) |
28 |
28 |
4 |
60 |
fluorite (kg) |
35 |
32 |
13 |
80 |
undersized ferroniobium (kg) |
25 |
25 |
-- |
50 |
total (kg) |
630 |
730 |
392 |
1752 |
amount of heat generation (kcal/kg) |
Δ 206 |
760 |
1590 |
600 |
[0040] Table 3 shows the composition of the resultant ferroniobium and the niobium yield
in the reaction. As apparent from the table, a high niobium yield of 97% was achieved.
It was also proved that the slag forming the upper layer of the product of a thermite
process according to the invention could be easily removed and the obtained ferroniobium
showed a very smooth surface.
[0041] For comparison, the composition of the produced ferroniobium and the niobium yield
in an experiment conducted by using a known thermite process are also shown in Table
3. A 1,000 kg of niobium ore was used to prepare a batch of a mixture of starting
materials (the contents being exactly same as those shown in the total column in Table
2), which was loaded at a time into a reactor furnace for thermite reaction. While
the reaction was terminated five minutes after the start of the reaction, it went
on very gradually in the initial stages and, after a few minutes, was accelerated
vehemently, fiercely spattering the materials to threaten the security of the operation.
The slag was less easily separated from the ferroniobium product than that of the
experiment conducted by way of a process according to the invention.
Table 3
|
|
produced ferroniobium (kg) |
composition (%) |
(Nb+Ta) yield (%) |
|
|
|
Nb+Ta |
Al |
Si |
Fe |
|
present invention |
1 |
649 |
68.1 |
1.3 |
2.1 |
bal. |
97.2 |
2 |
651 |
67.9 |
1.4 |
2.2 |
bal. |
97.1 |
3 |
645 |
68.3 |
1.0 |
2.0 |
bal. |
96.8 |
comparison |
4 |
611 |
66.7 |
1.4 |
2.5 |
bal. |
89.6 |
5 |
624 |
66.9 |
1.3 |
2.6 |
bal. |
91.8 |
[Example 2]
[0042] Metallic chromium was produced, using chromium oxide (purity: 99.9% in terms of Cr₂O₃)
as a starting metal oxide, aluminum (purity: 99.9%, particle size: 1 mm or less) as
a reducing agent, sodium chlorate (powder) as an exothermic agent.
[0043] A batch containing chromium oxide by 1,000 kg was divided into four lots, whose contents
are shown in Table 4, which also shows the calculated amount of heat generation for
each lot.
Table 4
|
1st lot |
2nd lot |
3rd lot |
4th lot |
total |
chromium oxide (kg) |
250 |
250 |
250 |
250 |
1000 |
sodium chlorate (kg) |
16.6 |
20.5 |
21.0 |
24.9 |
83 |
aluminum (kg) |
86.9 |
88.5 |
89.5 |
91.1 |
356 |
total (kg) |
353.5 |
359.0 |
360.5 |
366.0 |
1439 |
amount of heat generation (kcal/kg) |
685 |
715 |
723 |
751 |
719 |
[0044] For thermite reaction, a reactor furnace similar to the one used in Example 1 was
used and the four lots were sequentially loaded into the reactor furnace to form so
may layers. The top of the load was ignited to trigger a thermite reaction. The reaction
went on evenly until it was terminated three minutes after the start of the reaction.
When the reaction product was cooled, the slag found on the upper portion of the product
was removed and the metallic chromium, remained in the furnace was collected. It was
also proved that the slag forming the upper layer of the product could be easily removed.
[0045] For comparison, a batch whose composition was exactly same as that of the above examples
was loaded at a time into a reactor furnace for thermite reaction. While the reaction
was terminated fifty seconds after the start of the reaction, it went on very gradually
in the initial stages and, after tens of seconds, was accelerated vehemently, fiercely
spattering the materials. The slag was less easily separated from the metallic chromium
product than that of the experiment conducted by way of a process according to the
invention and the metallic chromium product contained slag to a considerable extent.
[0046] Table 5 shows the purity of the resultant metallic chromium and the chromium yield
of Example 2 as well as those of the experiment conducted for comparison. It is clearly
shown that a process according to the present invention can produce a higher purity
and a higher yield than a known method.
Table 5
|
produced metallic chromium |
(kg) composition (%) |
Cr yield (%) |
|
|
Cr |
Fe |
Si |
Al |
Cu |
C |
S |
|
the present invention |
(1) 575.6 |
99.93 |
0.0215 |
0.0065 |
0.0150 |
0.0010 |
0.0034 |
0.0216 |
84.1 |
(2) 574.9 |
99.93 |
0.0235 |
0.0058 |
0.0098 |
0.0010 |
0.0053 |
0.0232 |
84.0 |
comparison |
558.2 |
99.91 |
0.0230 |
0.0062 |
0.0318 |
0.0010 |
0.0042 |
0.0254 |
81.6 |
[Effects]
[0047] As is apparent from the above description, a thermite process for producing a metal
or alloy according to the invention ensures an even progress of thermite reaction
throughout the process, eliminating any unintended halt or excessive progress of reaction.
Consequently, it ensures a high purity as well as a high yield for metal or alloy
production.